|Publication number||US7751632 B2|
|Application number||US 11/058,558|
|Publication date||Jul 6, 2010|
|Priority date||Feb 15, 2005|
|Also published as||US20060182357|
|Publication number||058558, 11058558, US 7751632 B2, US 7751632B2, US-B2-7751632, US7751632 B2, US7751632B2|
|Inventors||Lipin Liu, K. C. Lee, Hasan Timucin OZDEMIR|
|Original Assignee||Panasonic Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (23), Non-Patent Citations (2), Referenced by (6), Classifications (11), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention generally relates to digital surveillance media storage systems, and relates in particular to intelligent selection and use of various extensible encoding formats to achieve efficient storage of surveillance media data based on analysis of media content and correlated events.
Surveillance applications usually require recording of all media (video/audio) for later access and retrieval for forensic purposes. Because of constant recording and need to keep archives for a long period of time, such as thirty days to three months or more, today's surveillance system needs either a large amount of videotapes for analog based system, or large amounts of disk space for digital recording and archive systems.
Conventional analog surveillance systems use analog signals to transmit, display, and record on analog-signal-based devices, and they store media archives in long-hour time-lapse tapes. A human expert is required to manage tapes and search and retrieve them manually. In contrast, a digital surveillance system converts analog signals to digital signals and encodes them to compress the data. Since encoding mechanisms are different, sizes of data vary according to the type of encoding process.
Digital surveillances are not only used in a variety application fields, but also have different needs relating to video content in terms of video encoding quality, storage efficiency, and others. Most digital surveillance systems offer only one, or very few, limited media encoding mechanisms.
Some problems of existing, conventional analog surveillance systems is that these systems do not enable convenient archive operations and are very time-consuming for searching and browsing a particular event video from storage archives.
Problems of existing, fixed-encoding-mechanism digital surveillance systems include: (a) a digital surveillance system records video using a fixed constant media rate regardless of the importance of a situation, thus consuming great amounts of storage space; (b) there are no content-based indexing and search systems to locate the target video efficiently; (c) a system provides only one type or very limited and pre-fixed types of media encoding methods; (d) there is no intelligence in encoding mechanism determination, such that manual operation is required to change the encoding type settings; and (e) the stored media is purged to tape or simply deleted if the disk space is full.
For long running operation, the disk space is always near full; therefore, the system needs to continue to purge or delete the media to make space available for new recordings.
An ideal digital surveillance system shall provide dynamic and scalable encoding and indexing methods that are needed based on alarm/event situations and/or actual video contents. The present invention, termed herein an Intelligent, dynamic, long-term digital Surveillance Media Storage System (ISMS), is a long-term digital media storage system with intelligent decision making based on event and content semantics to dynamically select the encoding method for new recording based on application defined priority and accuracy requirements. At the same time, it also, dynamically selects and converts previously recorded media to a less accurate recording format based on time and relative priority between a set of candidate media files and thus, forms a multistage media data compression mechanism that can best reduce the total storage size iteratively. This method traces computing resources with the storage resource to reduce the total system cost. The ISMS system can analyze contents and metadata of media streams, can provide flexible, scalable, and self-adjustable encoding and indexing mechanisms based on user defined rules, and can optimize the media storage in multiple stages according to time and semantically importance factors for digital surveillance systems. The system can include several subsystems, such as: (1) a media stream retriever/receiver; (2) a content-based description generator; (3) an intelligent engine; (4) an event handler; (5) a media stream writer; (6) a search and retrieval engine; (7) domain knowledge profiles; and (8) media file and metadata description index storage.
Further areas of applicability of the present invention will become apparent from the detailed description provided hereinafter. It should be understood that the detailed description and specific examples, while indicating the preferred embodiment of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
The present invention will become more fully understood from the detailed description and the accompanying drawings, wherein:
The following description of the preferred embodiments is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses.
The intelligent engine 12 gets: (a) segmented media data chunks from media analyzer; (b) metadata and event analysis from event metadata handler; and (c) content description metadata from description generator 18. These inputs will continue feeding into intelligent engine 12 during the analyzing process. The intelligent engine 12 processes inputs, checks rule conditions to determine what types of encoding mechanisms and what kinds of media quality should be used. Table 1 provides some examples of possible conditions and encoding modes.
The encoded data of segments are stored in data buffers until the whole media has been analyzed. Intelligent engine then instructs DSMF multiplexer and file writer 22 to create the DSMF media files and segmented media files, and store them in the media storage 24.
Suggested Encoding Mode
No (motion) change
Low-frame rate or skipping
frame still image encoding
Low motion objects
Still background with Object
High motion video
Full motion frame-based
High level alert
High Alert or
Full motion high quality
Most of surveillance applications perform twenty-four hours, seven days a week type of monitoring and recording, and the system generates large amounts of recording archives. Such systems can take advantage of intelligent dynamic encoding methods and storage optimization via multiple stage refinements to reduce the storage demands. As time goes by, most media archive files become less important unless there are events associated with the files or the semantic of the file contents are important. ISMS can re-encode media archives based on these factors to optimize the media storage. Although some of the applications, such as banks, may need to store the original video clips without cuts (for no motion condition), these applications can still gain the advantages from other types of encoding mechanisms and content analysis process with content-based indexing for fast search.
When the data is old and ready to be copied to other, off-line, secondary storage 72 devices such as tape or DVD-R, the system can still keep a reduced set of metadata and probably thumbnails to achieve fast retrieval based on metadata such as time, event, place, face, color, shape, and other content characteristics. The amount of data kept in metadata form is many orders of magnitude smaller than the compressed media. Therefore, the metadata can support online browsing. When a decision is made after the online browsing, specific tapes can be retrieved for further investigation. As result, this system can provide a virtual, semantic-based search interface for a very large amount of media data even if the data has been transferred to the secondary storage.
Moreover, a user-interface 70 of the ISMS system can permit a user to select media quality and encoding modes, define and select customized conditions, and/or select predefined conditions. The user interface 70 can also allow a user to set relationships between selected media quality, selected encoding modes, and selected conditions. These configurations can be stored in domain knowledge profiles 20. Further, the user interface 70 can allow a user to retrieve media streams from media file storage 24A using metadata description index 24B via search and retrieval engine 28 and media stream retriever 10B.
The domain knowledge profiles 20 contain: (a) rules 20B for selecting an encoding mode, which can be configured based on the actual needs of the applications; (b) the selectable encoder modules 20A which can be extended by importing a new encoder and registering it in the system; and (c) metadata description scheme 20C. The system can use either built-in or extensible configuration definition sets, called domain knowledge profiles, containing settings and rules to determine the best actions for different application domains and real situations. Encoding optimizer 42 follows the rules 20B defined in the profiles to make decisions to select stored files for optimization based on time and importance. Thus, the system can dynamically and intelligently select a proper encoding mechanism to record or transcode the video to long-term surveillance media storages. As a result, this flexible profile architecture allows extending the system to fit any application needs.
Functionalities of the system include digital surveillance information collection, media information processing, automated media description generation, intelligent encoding method selection, and media playback. Turning to
Encoding optimizer 42 periodically check if there is media file for optimization based on rules defined in profiles. If there is a file that needs to be optimized as at 41, encoding optimizer 42 retrieves the file 44 from media storage datastore 24 and sends it to media stream receiver 10 to start another stage of the encoding process for better optimization.
Turning now to
The format 48 contains the following: (a) media creation information 50; (b) media content that has been divided into multiple video segments (scenes or shots) 52; (c) media segmentation 54A with media encoding 54B, media type, time 54C, and media storage location 54D; and (d) content description of media segmentation.
Turning now to
The encoding modules 20A in the system are configurable based on the application and user needs. For example, the user can install new encoder modules and define the rules to use any of these modules. An encoding module can use a particular encoder with different profile@level which means different pre-defined quality, sampling rates, and resolutions provided by the encoder.
An unchanged frame-based encoding format can be employed as an initial phase of the skipping frames encoding format. Accordingly, an unchanged portion of an image is stored in relatively high resolution as relatively few frames over time that represent a background. A background frame is updated regularly to adapt to changing background conditions.
The skipping frame encoding 38A detects if the changes, including motion activity and color, of image frames is less than a pre-defined threshold. If so, the frames of the segment can be skipped as repeated, unchanged frames, thus keeping only certain amounts of full image frames (or still images) with time stamp metadata in it for video reconstruction. For example, the media can be received in the compression domain, with base frames and motion vectors for gradually changing the base frame (or an enhanced frame) into a next base frame (or enhanced frame). In this case, the motion vectors between base frames can be analyzed to determine if sufficiently rapid change is occurring based on the predefined change threshold. If not, then frames can be skipped. However, occasional base frames can be preserved and even enhanced with enhancement layers at a predetermined rate that can be set by a user in a related profile. In this way, slow changes, such as changes in lighting conditions, can be recorded, even if those changes are not rapid enough to exceed the predefined threshold.
It should be readily understood that the media analyzer of the present invention can semantically analyze these base frames (or enhanced frames) before the skipping frames encoder is selected. Accordingly, these base frames can be analyzed to determine if a complete lack of light condition has been reached, which can result in an alarm. Similarly, a slow moving robot under control of an interloper can be detected, resulting in an alarm. As a result, an interloper cannot foil the security system by slowly lowering lighting conditions until the cameras are useless, or by using a very slow, heat-controlled robot to avoid motion detection.
Turning now to
The present invention can provide the following advantages for a long-term digital surveillance media storage system: (a) reduce a great amount of long-term digital surveillance media storage space; (b) efficiently provide fast search capability for particular clips (e.g., as for evidence) from large amounts of long-term surveillance video storage; (c) provide intelligent, dynamic; and scalable recording quality control management; (d) acquire media content description data during analyzing and encoding processes; and (e) provide a content description for easy search and better understanding of media clips. The invention can be used with most digital surveillance systems that require storing media (video and/or audio) contents in recorders and long-term storage systems.
The description of the invention is merely exemplary in nature and, thus, variations that do not depart from the gist of the invention are intended to be within the scope of the invention. Such variations are not to be regarded as a departure from the spirit and scope of the invention.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5272529 *||Mar 20, 1992||Dec 21, 1993||Northwest Starscan Limited Partnership||Adaptive hierarchical subband vector quantization encoder|
|US5602594 *||Aug 12, 1994||Feb 11, 1997||Goldstar Co., Ltd.||Buffer control apparatus for compression of video intraframes|
|US5657087 *||Jun 15, 1995||Aug 12, 1997||Samsung Electronics Co., Ltd.||Motion compensation encoding method and apparatus adaptive to motion amount|
|US5894333 *||Jun 27, 1996||Apr 13, 1999||Mitsubishi Denki Kabushiki Kaisha||Representative image display method, representative image display apparatus, and motion image search appratus employing the representative image display apparatus|
|US6597738 *||Aug 27, 1999||Jul 22, 2003||Hyundai Curitel, Inc.||Motion descriptor generating apparatus by using accumulated motion histogram and a method therefor|
|US6671454 *||Nov 19, 1999||Dec 30, 2003||Nec Corporation||Program information providing apparatus and record/playback control apparatus|
|US6681058 *||Mar 6, 2000||Jan 20, 2004||Sarnoff Corporation||Method and apparatus for estimating feature values in a region of a sequence of images|
|US7069342 *||Mar 1, 2001||Jun 27, 2006||Cisco Technology, Inc.||Communication system with content-based data compression|
|US7177355 *||Dec 1, 1998||Feb 13, 2007||Samsung Electronics Co., Ltd.||Method for representing encoding upon coding information|
|US20010021268 *||Mar 6, 2001||Sep 13, 2001||Lg Electronics Inc.||Hierarchical hybrid shot change detection method for MPEG-compressed video|
|US20010036271 *||Mar 20, 2001||Nov 1, 2001||Javed Shoeb M.||System and method for securely distributing digital content for short term use|
|US20020114525 *||Feb 21, 2001||Aug 22, 2002||International Business Machines Corporation||Business method for selectable semantic codec pairs for very low data-rate video transmission|
|US20030154344 *||Feb 7, 2003||Aug 14, 2003||Nec Corporation||Data recording and reproducing method and apparatus capable of recompressing individual data blocks|
|US20030210821 *||Apr 20, 2001||Nov 13, 2003||Front Porch Digital Inc.||Methods and apparatus for generating, including and using information relating to archived audio/video data|
|US20040042639||Sep 16, 2003||Mar 4, 2004||Vladimir Pavlovic||Method for motion classification using switching linear dynamic system models|
|US20040143602 *||Oct 17, 2003||Jul 22, 2004||Antonio Ruiz||Apparatus, system and method for automated and adaptive digital image/video surveillance for events and configurations using a rich multimedia relational database|
|US20050058431 *||Sep 12, 2003||Mar 17, 2005||Charles Jia||Generating animated image file from video data file frames|
|US20050099498 *||Nov 7, 2003||May 12, 2005||Ich-Kien Lao||Digital video system-intelligent information management system|
|US20050117636 *||Oct 1, 2004||Jun 2, 2005||Aron Rosenberg||Methods and apparatus for encoding and decoding video data|
|US20050123048 *||Nov 5, 2004||Jun 9, 2005||Seiko Epson Corporation||Image data compression device and encoder|
|US20050229234 *||Jun 4, 2003||Oct 13, 2005||Koninklijke Philips Electronics N.V.||Method of filtering a bitstream according to user specifications|
|US20060245656 *||Nov 27, 2003||Nov 2, 2006||Junko Shinohara||Image encoding device and image encoding method|
|WO2001082255A1 *||Apr 24, 2000||Nov 1, 2001||Video Domain Technologies Ltd.||Surveillance system with camera|
|1||Introduction to MPEG-7 (v4.0), International Oganisation for Standardisation Organisation Internationale De Normalisation ISO/IEC JTC1/SC29/WG11, Mar. 2002.|
|2||*||Tang et al. ("Histological Image Retrieval Based on Semantic Content Analysis," IEEE Trans. Information Technology in Biomedicine, vol. 7, No. 1, Mar. 2003, pp. 26-36).|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US9041812 *||Nov 13, 2012||May 26, 2015||International Business Machines Corporation||Automated authorization to access surveillance video based on pre-specified events|
|US9071807||Jan 25, 2013||Jun 30, 2015||International Business Machines Corporation||Providing emergency access to surveillance video|
|US9191632 *||Jan 25, 2013||Nov 17, 2015||International Business Machines Corporation||Automated authorization to access surveillance video based on pre-specified events|
|US20120096007 *||Apr 19, 2012||Microsoft Corporation||Embedding content-based searchable indexes in multimedia files|
|US20140132765 *||Jan 25, 2013||May 15, 2014||International Business Machines Corporation||Automated Authorization to Access Surveillance Video Based on Pre-Specified Events|
|US20140132772 *||Nov 13, 2012||May 15, 2014||International Business Machines Corporation||Automated Authorization to Access Surveillance Video Based on Pre-Specified Events|
|U.S. Classification||382/239, 382/236, 382/305|
|International Classification||G06K9/46, G06K9/36, G06K9/60, G06K9/54|
|Cooperative Classification||G06K9/00771, G08B13/19669|
|European Classification||G08B13/196S2, G06K9/00V4|
|Feb 15, 2005||AS||Assignment|
Owner name: MATSUSHITA ELECTRIC INDUSTRIAL CO., LTD., JAPAN
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:LIU, LIPIN;LEE, K. C.;OZDEMIR, HASAN TIMUCIN;REEL/FRAME:016280/0233
Effective date: 20050214
|Nov 24, 2008||AS||Assignment|
Owner name: PANASONIC CORPORATION, JAPAN
Free format text: CHANGE OF NAME;ASSIGNOR:MATSUSHITA ELECTRIC INDUSTRIAL CO., LTD.;REEL/FRAME:021897/0707
Effective date: 20081001
Owner name: PANASONIC CORPORATION,JAPAN
Free format text: CHANGE OF NAME;ASSIGNOR:MATSUSHITA ELECTRIC INDUSTRIAL CO., LTD.;REEL/FRAME:021897/0707
Effective date: 20081001
|Nov 2, 2010||CC||Certificate of correction|
|Dec 11, 2013||FPAY||Fee payment|
Year of fee payment: 4